Postoperative red blood cell transfusion and morbid outcome in uncomplicated cardiac surgery patients.

OBJECTIVE: To evaluate postoperative red blood cell (RBC) transfusion and its association with postoperative cardiac events and multiorgan morbidity in uncomplicated cardiac surgery patients. METHODS: A cohort of 945 patients from the 5,436 coronary artery bypass grafting patients enrolled in the international Multicenter Study of Perioperative Ischemia (McSPI) Epidemiology II (EPI II) study was investigated. Inclusion criteria were low to moderate risk profile, postoperative hemoglobin level ≥ 10 g/dl, minimal postoperative blood loss, and no evidence of any morbid event on the day of surgery. RBC transfusion was assessed during the first 24 postoperative hours and cardiac as well as multiorgan outcomes from postoperative day 2 to hospital discharge. Multivariate analysis was applied to assess the effect of RBC transfusion on multiorgan outcomes. A secondary propensity score analysis was performed in 4,465 patients without early postoperative morbid outcomes. RESULTS: Transfused patients (193/945, 20.4%) were more likely to suffer cardiac events (P = 0.03), harvest-site infection (P = 0.002), and composite morbidity outcome (P = 0.04). RBC transfusion was associated with cardiac events on multivariate as well as on propensity score analysis (adjusted odds ratio, 1.39; 95% confidence interval, 1.01-1.92; P = 0.04), and with harvest-site infection on multivariate analysis. Additionally, propensity score analysis suggested possible associations of RBC transfusion with increased risks for composite morbidity outcome and in-hospital mortality, renal morbidity, pneumonia, and mediastinitis. CONCLUSIONS: The data suggest a potential association between postoperative RBC transfusion and increased morbidity for cardiac surgery patients with low to moderate mortality risk profiles, adequate hemoglobin levels, and low bleeding rates.

The ongoing variability in blood transfusion practices in cardiac surgery.

BACKGROUND: Although blood utilization has been under considerable scrutiny for the past two decades, particularly for surgery, studies comparing perioperative blood transfusion practices between countries are rare, and the evolution of international standards remains unknown. Therefore, the objective of this evaluation was to compare the perioperative transfusion of blood components in cardiac surgery in multiple centers in different countries. STUDY DESIGN AND METHODS: Transfusion practice was investigated prospectively in 70 centers among 16 countries. A total of 5065 randomly selected cardiac surgery patients of the Multicenter Study of Perioperative Ischemia Epidemiology II (EPI II) Study were evaluated. Utilization of red blood cells (RBCs), fresh-frozen plasma (FFP), and platelets (PLTs) was assessed daily, before, during, and after surgery until hospital discharge. RESULTS: Intraoperative RBC transfusion varied from 9 to 100 percent among the 16 countries, and 25 to 87 percent postoperatively (percentage of transfused patients). Similarly, frequency of transfusion of FFP varied from 0 to 98 percent intraoperatively and 3 to 95 percent postoperatively, and PLT transfusion from 0 to 51 and 0 to 39 percent, respectively. Moreover, there were not only marked differences in transfusion rates between centers in different countries but also in interinstitutional comparison of multiple centers within countries. CONCLUSION: In cardiac surgical patients, marked variability in transfusion practice exists between centers in various countries and suggests differences in perioperative practice patterns as well as possible inappropriate use. International standardization of perioperative practice patterns as well as transfusion regimes appears necessary.

Impact of preoperative anemia on outcome in patients undergoing coronary artery bypass graft surgery.

BACKGROUND: The risk of preoperative anemia in patients undergoing heart surgery has not been described precisely. Specifically, the impact of low hemoglobin per se or combined with other risk factors on postoperative outcome is unknown. Thus, we determined the effects of low preoperative hemoglobin and comorbidities on postoperative adverse outcomes in patients with coronary artery bypass graft in a large comprehensive multicenter study. METHODS AND RESULTS: The Multicenter Study of Perioperative Ischemia investigated 5065 patients with coronary artery bypass graft at 70 institutions worldwide, collecting approximately 7500 data points per patient. In 4804 patients who received no preoperative transfusions, we determined the association between lowest preoperative hemoglobin levels and in-hospital cardiac and noncardiac morbidity and mortality and the impact of concomitant risk factors, assessed by EuroSCORE, on this effect. In patients with EuroSCORE < 4 (n=2054), only noncardiac outcomes were increased, whereas patients with EuroSCORE > or = 4 (n=2750) showed an increased incidence of all postoperative events, starting at hemoglobin < 11 g/dL. Low preoperative hemoglobin was an independent predictor for noncardiac (renal > cerebral; P<0.001) outcomes, whereas the increase in cardiac events was due to other factors associated with preoperative anemia. CONCLUSIONS: Anemic patients undergoing cardiac surgery have an increased risk of postoperative adverse events. Importantly, the extent of preexisting comorbidities substantially affects perioperative anemia tolerance. Therefore, preoperative risk assessment and subsequent therapeutic strategies, such as blood transfusion, should take into account both the individual level of preoperative hemoglobin and the extent of concomitant risk factors.

Coronary artery bypass graft surgery--care globalization: the impact of national care on fatal and nonfatal outcome.

OBJECTIVE: In an international, prospective, observational study, we contrasted adverse vascular outcomes among four countries and then assessed practice pattern differences that may have contributed to these outcomes. METHODS: A total of 5065 patients undergoing coronary artery bypass graft surgery were analyzed at 70 international medical centers, and from this pool, 3180 patients from the 4 highest enrolling countries were selected. Fatal and nonfatal postoperative ischemic complications related to the heart, brain, kidney, and gastrointestinal tract were assessed by blinded investigators. RESULTS: In-hospital mortality was 1.5% (9/619) in the United Kingdom, 2.0% (9/444) in Canada, 2.7% (34/1283) in the United States, and 3.8% (32/834) in Germany (P = .03). The rates of the composite outcome (morbidity and mortality) were 12% in the United Kingdom, 16% in Canada, 18% in the United States, and 24% in Germany (P < .001). After adjustment for difference in case-mix (using the European System for Cardiac Operative Risk Evaluation) and practice, country was not an independent predictor for mortality. However, there was an independent effect of country on composite outcome. The practices that were associated with adverse outcomes were the intraoperative use of aprotinin, intraoperative transfusion of fresh-frozen plasma or platelets, lack of use of early postoperative aspirin, and use of postoperative heparin. CONCLUSIONS: Significant between-country differences in perioperative outcome exist and appear to be related to hematologic practices, including administration of antifibrinolytics, fresh-frozen plasma, platelets, heparin, and aspirin. Understanding the mechanisms for these observations and selection of practices associated with improved outcomes may result in significant patient benefit.

Molecular markers of brain damage--clinical and ethical implications with particular focus on cardiac arrest.

Although 25-50% of patients suffering from cardiac arrest can be stabilised haemodynamically, the hospital discharge rate is only 2-14%. One of the major causes of this discrepancy is persistent brain damage. Studies to assess the prognostic value of early prediction of neurologic and overall outcome in patients with cardiac arrest have not yet produced precise and generally accepted diagnostic rules. As apparative diagnostic methods often fail to predict neurologic outcome, the role of molecular markers has come a focus of common interest for early outcome prediction. This systematic review article aims to give an overview on the most important molecular markers for neurologic and overall outcome prediction and outline the advantages, clinical implications and ethical issues in patients undergoing cardiopulmonary resuscitation after cardiac arrest. For this purpose, the traditional marker for brain damage, the neuron-specific enolase, a gamma gamma isomer of enolase and cytoplasmatic enzyme of glycolysis, and the astroglial protein S100, a calcium-binding protein regulating neuronal differentiation, outgrowth, and apoptosis, are analysed and their role discussed as a marker for brain damage in general and recovery after cardiopulmonary resuscitation following cardiac arrest. Neuron-specific enolase has been investigated as a neuro-marker after brain damage and for outcome prediction in unconscious patients. Whereas the protein S100 has proven to be a good marker for neuronal damage after isolated brain injury, its role in cardiac surgery is not as clear: at least, in the early postoperative phase S100 is not a sole marker for neurologic damage, as release of S100 from cardiac tissue and other sources has also been demonstrated. However, the persistent elevation of S100 after cardiac surgery is specific for neurologic impairment. Most interestingly, after cardiac arrest the protein S100 has shown to be a good survival marker for overall outcome prediction. Although it cannot be absolutely determined whether cerebral or cardiac release of S100 is predominant in this clinical setting, recent studies have revealed that S100 serum levels are a useful diagnostic tool for outcome prediction. In contrast, after cardiac arrest serum levels of protein S100 did not reach a 100% specificity and sensitivity in clinical studies, and, therefore, elevated S100 in these patients has to be interpreted with caution. Nonetheless, low S100 serum levels have been correlated with good outcome and, therefore, even if all other diagnostic tests indicate poor outcome, all therapeutic efforts must be undertaken, as no single study has shown that normal S100 serum levels were associated with poor prognosis.